OPEC and civil unrest

[Camarohusky]

OPEC has been accused of inflating oil prices this entire decade. Becuase of the dramatic rise in oil prices the price of food has skyrocketed. Earlier this year we saw three countries fall into unrest because of skyrocketing food prices. OPEC countries are mostly stiff dictatorships with large impoverished populations/

So I put to you, will OPEC's inflation of oil prices end up in the leaders being overthrown? Will the massive rise in the cost of good globally lead to people in Venzuela, Saudi Arabia, Niger, and the other OPEC countries?

[Iron-Hampster]

Saudi Arabia seems too untouched by these uprisings. If we wan't to put OPEC in it's place, we need to either:

Solution 1: produce much more oil, and under sell them

Solution 2: find an alternative fuel that can replace petroleum in our combustion engines (ex: hydrogen) or completely ditch combustion engines in favor of a completely different mechanic. (ex: electric)

[Der-Lowe]

Okay, let's go by parts. What I think you are saying can be brought down to the following hypothesis and conclusions:

1) OPEC inflates oil prices through the restraining of their production. Ceteris paribus, if they are willing to supply less, there is an oil shortage, and demand must fall through an increase on prices.

Graphically, the demand is unchanged, and the OPEC moves along the demand curve to maximize profit.

We should then see the following in the data for the last years:
a) Rising prices
b) Falling OPEC production
And since (I think) you believe OPEC only is colluding
c) An increase in production from the non OPEC countries, since they behave competitively and increase supply when price increases.

Let me say before getting and processing the data, that your position is more restrictive than what one could originally think. What it is implied in your reasoning is that not only OPEC colludes, but also, that it started colluding now, because otherwise, we'd see the production rationing in the past, and price and production would behave in the same way than in a competitive scenario (price and demand go in the same direction).

My alternative hypothesis to this point is (and was before doing the research) that the price increase was not because of non competitive tactics, but rather by an increase in demand. If my hypothesis is true, then we'd see:

a) Rising prices
b) Increased production both in OPEC and non OPEC nations.

My hypothesis also requires an additional assumption, that oil production shows decreasing returns (supply curve to exist and have a positive slope).

Data
Methodological approach: The two variables that interest us are price of oil, and quantity produced, both by OPEC and by non OPEC nations. For the price of oil I have chosen the U.S. Crude Oil Imported Acquisition Cost by Refiners (Dollars per Barrel) from the Energy Information Administration. My reason for choosing this specific variable is that it is what it is considered to be the "oil world price" since it is a weighted average of all the oil the US buys from every corner of the planet, and data availability. I have brought it to real terms by adjusting with the US CPI
Finally OPEC production data is available from BP Global
I have examined the 2000-9 period on an annual basis.
I will propose the following models

OPECi = a1+b1 * Pricei+ui ; i=1,...,10
NotOPECi = a2+b2 * Pricei+ui ; i=1,...,10

Where OPECi is the aggregate crude oil production of opec nations for the period i in thousands of barrels a day,
NotOPEC is aggregate crude oil production of nations not in the OPEC for the period i n thousands of barrels a day,
Price is the U.S. Crude Oil Imported Acquisition Cost by Refiners for period i in Real Dollars per Barrel.

and solve a1,a2,b1,b2 through the least squares estimation method. Therefore, I will be using classical assumptions.

Results

The stata log is as following:

. regress notopec price if year>1999

Source | SS df MS Number of obs = 10
-------------+--------------------------
---- F( 1, 8) = 30.23
Model | 96704780.3 1 96704780.3 Prob > F = 0.0006
Residual | 25588457.6 8 3198557.2 R-squared = 0.7908
-------------+--------------------------
---- Adj R-squared = 0.7646
Total | 122293238 9 13588137.5 Root MSE = 1788.5

----------------------------------------
--------------------------------------
notopec | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+--------------------------
--------------------------------------
price | 329.7299 59.96687 5.50 0.001 191.446 468.0137
_cons | 25713.65 1507.68 17.06 0.000 22236.94 29190.37
----------------------------------------
--------------------------------------

. regress opec price if year>1999

Source | SS df MS Number of obs = 10
-------------+--------------------------
---- F( 1, 8) = 25.73
Model | 34350143.7 1 34350143.7 Prob > F = 0.0010
Residual | 10681420.7 8 1335177.59 R-squared = 0.7628
-------------+--------------------------
---- Adj R-squared = 0.7332
Total | 45031564.5 9 5003507.16 Root MSE = 1155.5

----------------------------------------
--------------------------------------
opec | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+--------------------------
--------------------------------------
price | 196.5163 38.74395 5.07 0.001 107.1726 285.86
_cons | 28230.74 974.0958 28.98 0.000 25984.47 30477.01
----------------------------------------
--------------------------------------

That is, we see a direct relationship between price and not opec, and between price and opec.
More precisely, an increase of the price of oil in one real dollar increases opec production by 196.51 thousand barrels a day in average, and not opec production in 329.73
Despite the small sample, the probability of price and production having no relationship is very low 0.1%, whereas the probability of price reducing production is even smaller, half of it (being the t distribution symmetric), 0.05%.

And here's the graph: